Docker Cheat Sheet

NOTE: This used to be a gist that continually expanded. It's now a github project because it's considerably easier for other people to edit, fix and expand on Docker using Github. Just click README.md, and then on the "writing pen" icon on the right to edit.

Why

"With Docker, developers can build any app in any language using any toolchain. “Dockerized” apps are completely portable and can run anywhere - colleagues’ OS X and Windows laptops, QA servers running Ubuntu in the cloud, and production data center VMs running Red Hat.

Developers can get going quickly by starting with one of the 13,000+ apps available on Docker Hub. Docker manages and tracks changes and dependencies, making it easier for sysadmins to understand how the apps that developers build work. And with Docker Hub, developers can automate their build pipeline and share artifacts with collaborators through public or private repositories.

Docker helps developers build and ship higher-quality applications, faster." -- What is Docker

Prerequisites

I use Oh My Zsh with the Docker plugin for autocompletion of docker commands. YMMV.

Linux

The 3.10.x kernel is the minimum requirement for Docker.

MacOS

Use Homebrew.

ruby -e "$(curl -fsSL https://raw.github.com/mxcl/homebrew/go)"

Installation

Installation instructions are available. Docker has recognized that installation / deployment is a PITA, and Docker Machine aka bug 8681 deals with this specifically.

Linux

Quick and easy install script provided by Docker:

curl -sSL https://get.docker.com/ | sh

If you're not willing to run a random shell script, please see the installation instructions for your distribution.

Mac OS X

Download Docker for OSX from the Github Releases page.

The canonical way to use Docker is with the aid of the boot2docker VM. However, using the out of the box boot2docker doesn't give me control over my Vagrant instances (especially the lack of port forwarding). So here's how to use boot2docker from a Vagrant instance.

We use the YungSang modified boot2docker instance from the Vagrant Cloud:

mkdir ~/boot2docker
cd ~/boot2docker
vagrant init yungsang/boot2docker
vagrant up
export DOCKER_HOST=tcp://localhost:2375
docker version

NOTE: the YungSang boot2docker opens up port forwarding to the network, so is not safe on public wifi. You can make a good argument that docker without TLS is fundamentally unsafe. I only do it because I have Hands Off installed to limit external network access.

Then start up a container:

docker run -i -t ubuntu /bin/bash

That's it, you have a running Docker container.

Containers

Your basic isolated Docker process. Containers are to Virtual Machines as threads are to processes. Or you can think of them as chroots on steroids.

Lifecycle

If you want to run and then interact with a container, docker start, then spawn a shell as described in Executing Commands.

If you want a transient container, docker run --rm will remove the container after it stops.

If you want to remove also the volumes associated with the container, the deletion of the container must include the -v switch like in docker --rm -v.

If you want to poke around in an image, docker run -t -i <myimage> <myshell> to open a tty.

If you want to poke around in a running container, docker exec -t -i <mycontainer> <myshell> to open a tty.

If you want to map a directory on the host to a docker container, docker run -v $HOSTDIR:$DOCKERDIR. Also see Volumes.

If you want to integrate a container with a host process manager, start the daemon with -r=false then use docker start -a.

If you want to expose container ports through the host, see the exposing ports section.

Restart policies on crashed docker instances are covered here.

Info

docker ps -a shows running and stopped containers.

Import / Export

  • docker cp copies files or folders between a container and the local filesystem..
  • docker export turns container filesystem into tarball archive stream to STDOUT.

Executing Commands

To enter a running container, attach a new shell process to a running container called foo, use: docker exec -it foo /bin/bash.

Images

Images are just templates for docker containers.

Lifecycle

Info

Docker image ids are sensitive information and should not be exposed to the outside world. Treat them like passwords.

Registry & Repository

A repository is a hosted collection of tagged images that together create the file system for a container.

A registry is a host -- a server that stores repositories and provides an HTTP API for managing the uploading and downloading of repositories.

Docker.com hosts its own index to a central registry which contains a large number of repositories. Having said that, the central docker registry does not do a good job of verifying images and should be avoided if you're worried about security.

Run local registry

Registry implementation has an official image for basic setup that can be launched with docker run -p 5000:5000 registry Note that this installation does not have any authorization controls. You may use option -P -p 127.0.0.1:5000:5000 to limit connections to localhost only. In order to push to this repository tag image with repositoryHostName:5000/imageName then push this tag.

Dockerfile

The configuration file. Sets up a Docker container when you run docker build on it. Vastly preferable to docker commit. If you use jEdit, I've put up a syntax highlighting module for Dockerfile you can use. You may also like to try the tools section.

Instructions

Tutorial

Examples

Best Practices

This is where general Docker best practices and war stories go:

Layers

The versioned filesystem in Docker is based on layers. They're like git commits or changesets for filesystems.

Note that if you're using aufs as your filesystem, Docker does not always remove data volumes containers layers when you delete a container! See PR 8484 for more details.

Links

Links are how Docker containers talk to each other through TCP/IP ports. Linking into Redis and Atlassian show worked examples. You can also (in 0.11) resolve links by hostname.

NOTE: If you want containers to ONLY communicate with each other through links, start the docker daemon with -icc=false to disable inter process communication.

If you have a container with the name CONTAINER (specified by docker run --name CONTAINER) and in the Dockerfile, it has an exposed port:

EXPOSE 1337

Then if we create another container called LINKED like so:

docker run -d --link CONTAINER:ALIAS --name LINKED user/wordpress

Then the exposed ports and aliases of CONTAINER will show up in LINKED with the following environment variables:

$ALIAS_PORT_1337_TCP_PORT
$ALIAS_PORT_1337_TCP_ADDR

And you can connect to it that way.

To delete links, use docker rm --link .

If you want to link across docker hosts then you should look at Swarm. This link on stackoverflow provides some good information on different patterns for linking containers across docker hosts.

Volumes

Docker volumes are free-floating filesystems. They don't have to be connected to a particular container. You should use volumes mounted from data-only containers for portability.

Volumes are useful in situations where you can't use links (which are TCP/IP only). For instance, if you need to have two docker instances communicate by leaving stuff on the filesystem.

You can mount them in several docker containers at once, using docker run --volumes-from.

Because volumes are isolated filesystems, they are often used to store state from computations between transient containers. That is, you can have a stateless and transient container run from a recipe, blow it away, and then have a second instance of the transient container pick up from where the last one left off.

See advanced volumes for more details. Container42 is also helpful.

For an easy way to clean abandoned volumes, see docker-cleanup-volumes

As of 1.3, you can map MacOS host directories as docker volumes through boot2docker:

docker run -v /Users/wsargent/myapp/src:/src

You can also use remote NFS volumes if you're feeling brave.

You may also consider running data-only containers as described here to provide some data portability.

Exposing ports

Exposing incoming ports through the host container is fiddly but doable.

The fastest way is to map the container port to the host port (only using localhost interface) using -p:

docker run -p 127.0.0.1:$HOSTPORT:$CONTAINERPORT --name CONTAINER -t someimage

If you don't want to use the -p option on the command line, you can persist port forwarding by using EXPOSE:

EXPOSE <CONTAINERPORT>

If you're running Docker in Virtualbox, you then need to forward the port there as well, using forwarded_port. It can be useful to define something in Vagrantfile to expose a range of ports so that you can dynamically map them:

Vagrant.configure(VAGRANTFILE_API_VERSION) do |config|
  ...

  (49000..49900).each do |port|
    config.vm.network :forwarded_port, :host => port, :guest => port
  end

  ...
end

If you forget what you mapped the port to on the host container, use docker port to show it:

docker port CONTAINER $CONTAINERPORT

Machines, Swarm and Compose

In early 2015 Docker developed and open sourced a number of components that use the Docker API to provide some higher level services.

  • Machine, is a tool for easily creating a docker host either locally or on cloud provider (i.e. AWS, Digital Ocean etc).
  • Swarm, is clustering for docker. It provides a way to group a number of docker hosts into a single virtual entity and provides mechanisms to schedule containers across these hosts.
  • Compose, provides a tool for describing a multi-container application within a single file. Compose primarily Fig renamed.

The Demo of the Machine + Swarm + Compose integration video provides a good introduction to the various components and how they work together.

Tips

Sources:

Last Ids

alias dl='docker ps -l -q'
docker run ubuntu echo hello world
docker commit `dl` helloworld

Commit with command (needs Dockerfile)

docker commit -run='{"Cmd":["postgres", "-too -many -opts"]}' `dl` postgres

Get IP address

docker inspect `dl` | grep IPAddress | cut -d '"' -f 4

or

wget http://stedolan.github.io/jq/download/source/jq-1.3.tar.gz
tar xzvf jq-1.3.tar.gz
cd jq-1.3
./configure && make && sudo make install
docker inspect `dl` | jq -r '.[0].NetworkSettings.IPAddress'

or using a go template

docker inspect -f '{{ .NetworkSettings.IPAddress }}' <container_name>

Get port mapping

docker inspect -f '{{range $p, $conf := .NetworkSettings.Ports}} {{$p}} -> {{(index $conf 0).HostPort}} {{end}}' <containername>

Find containers by regular expression

for i in $(docker ps -a | grep "REGEXP_PATTERN" | cut -f1 -d" "); do echo $i; done`

Get Environment Settings

docker run --rm ubuntu env

Kill running containers

docker kill $(docker ps -q)

Delete old containers

docker ps -a | grep 'weeks ago' | awk '{print $1}' | xargs docker rm

Delete stopped containers

docker rm -v `docker ps -a -q -f status=exited`

Delete dangling images

docker rmi $(docker images -q -f dangling=true)

Delete all images

docker rmi $(docker images -q)

Show image dependencies

docker images -viz | dot -Tpng -o docker.png

Slimming down Docker containers Intercity Blog

  • Cleaning APT
RUN apt-get clean
RUN rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
  • Flatten an image
ID=$(docker run -d image-name /bin/bash)
docker export $ID | docker import – flat-image-name
  • For backup
ID=$(docker run -d image-name /bin/bash)
(docker export $ID | gzip -c > image.tgz)
gzip -dc image.tgz | docker import - flat-image-name

Monitor system resource utilization for running containers

To check the CPU, memory and network i/o usage, you can use:

docker stats <container>

for a single container or

docker stats $(docker ps -q)

to monitor all containers on the docker host.

Tools

Compose

Compose (originally named fig) is a helper app that makes it easier to run multiple docker containers on the same host. I would expect it to be used during dev/qa more than in production.

Compose works with a YAML file (default name can be docker-compose.yml, use -f to provide a different filename) that defines the containers you wish to use with it. Compose will take its project name from the name of the folder containing your yml configuration but you can override that with the -p parameter.

Once I have my config defined, I can use docker-compose up -d to run it (the -d runs it as a detached task). This will start and link any containers.

You can do everything you do with compose using docker directly but running multiple containers with parameters would require some sort of script if you plan to do it more than once so the yml config of compose and the convenience commands it provides are worth considering.

Here's an example of setting up a docker-compose.yml for an app with an apache packaged client container and a tomcat packaged app war:

First, here are the two docker commands to run these containers:

docker run -p 8080:8080 -v /Users/me/tomcatwork/trial.properties:/usr/share/tomcat6/trial.properties:rw -d me/tcfull
docker run -p 80:80 -v /Users/me/dockerwork/localproxy.conf:/etc/apache2/conf-enabled/proxy.conf:rw -d me/afull

At this point, I haven't linked the containers - I'm using the proxy.conf to specify the tomcat address. my docker-compose.yml looks like this:

app:
  image: me/tcfull
  ports:
    - "8080"
  volumes:
    - /Users/me/tomcatwork/trial.properties:/usr/share/tomcat6/trial.properties:rw
web:
  image: me/afull
  ports:
    - "80:80"
  volumes:
    - /Users/me/figwork/proxy.conf:/etc/apache2/conf-enabled/proxy.conf:rw
  links:
    - app

As you can see it follows the docker commands with the addition of names for the containers and a links section for the web container, linking it to the app container.

As part of that linking process, docker will copy any environment variables defined in the app container over to the web container, define new environment variables for the address the app container is running at and also add an app entry in the etc/hosts file for the web container. I can modify my proxy conf to address http://app:8080 and fig/docker will take care of the rest.

I can then use commands like docker-compose stop and docker-compose rm to stop all my containers and remove them. NB - docker will eventually absorb figs functionality with docker groups and docker up but it looks like they're keeping the yml config so it should be pretty seamless when it happens.

Troubleshooting

docker-compose run is a useful command for debugging issues. It allows me to startup a named container (and any it links to) and run a one off command.

This allows me to do things like docker-compose run web env which will give me a list of all the environment variables that are available on the web container including the ones generated via the link to app.

I can also use docker-compose run web bash to run my web container interactively the way it has been setup by compose with the link to app so I can debug any issues from the command line.

Panamax

Nice web UI, will let you set up and download multiple docker containers.

Vessel

Vessel automates the setup & use of dockerized development environments.